I turn AI ideas into systems that survive production.

ABC Teknology

• Designed and developed a production AI shopping assistant for UAE grocery price comparison, using a FastAPI backend and LangGraph agent architecture to interpret natural-language shopping requests, search retailers in parallel, filter results, and rank products by relevance and value.
• Built parallel web-crawling and data-ingestion services for major UAE grocery retailers, supporting live product, price, availability, and catalog data for AI-powered shopping workflows.
• Developed a cross-platform React Native / Expo mobile application for iOS and Android with TypeScript, multi-provider OAuth, secure session management, and real-time backend communication.
• Extended the same AI backend to WhatsApp, enabling users to interact with the shopping assistant through multiple channels using Twilio.
• Implemented LLM-driven product ranking, availability filtering, relevance filtering, live progress streaming, and multi-step agent orchestration using LangGraph and OpenAI.
• Designed backend services with PostgreSQL and Redis for persistent data, caching, session management, and scalable request handling.
• Worked across AI architecture, backend APIs, web scraping, mobile development, authentication, integrations, and production deployment.

StechAI

• Designed and deployed a production RAG platform for enterprise knowledge retrieval across 100K+ documents, using LangChain, Qdrant, hybrid retrieval, citation grounding, FastAPI, Redis, GPT-4, and Claude.
• Implemented MCP-based AI agents capable of multi-step tool use and reasoning, extending the RAG platform beyond document retrieval into agentic workflows.
• Optimized document chunking, retrieval, caching, and validation pipelines, achieving 94% retrieval accuracy and approximately 70% faster search.
• Built TrendGen, an end-to-end AI trend intelligence platform processing 100K+ social posts per run using embeddings, Sentence Transformers, BERTopic, LLM-assisted trend naming, Qdrant, and MongoDB.
• Developed emerging-signal forecasting using Qdrant vector search and Dynamic Time Warping (DTW), achieving a 74% emerging-signal hit rate with approximately 17 days average forecast lead time.
• Built asynchronous ML job orchestration with live progress streaming, scheduled pipelines, model versioning, and Azure Blob Storage for repeatable production inference.
• Developed a React and TypeScript administration console for trend generation, model management, trend exploration, forecasting, monitoring, and analytics.
• Designed and deployed FastAPI services and data pipelines integrating MongoDB, Qdrant, Azure Storage, OpenAI, and scheduled ML workflows.

Freelance

• Designed and delivered production AI, automation, CRM, marketing, conversational AI, and full-stack products for international clients across service, sales, marketing, and technology use cases.
• Built Tap Connect, a production digital identity and networking platform using Next.js, FastAPI, PostgreSQL, Redis, HubSpot, NFC, and QR workflows, including live profiles, card management, connection capture, analytics, team governance, and CRM integrations.
• Developed Communication AI Hub, a production CRM and communications operations platform integrating lead intake, automated nurture flows, email, SMS, voice, call review, inbound AI agents, human review queues, and CRM workflows.
• Built Marketing AI Hub, a multi-brand content and advertising platform that transforms field videos into AI-generated social content, campaigns, organic posts, and Meta advertising workflows with human approval and brand controls.
• Designed and deployed AI voice agents and omnichannel booking systems using VAPI, OpenAI, Twilio, FastAPI, n8n, and Airtable for automated phone, SMS, and web booking.
• Developed intelligent document-processing and knowledge systems using RAG, vector databases, LLMs, and MCP-based agent workflows for automated extraction, retrieval, validation, and reporting.
• Built full-stack AI applications combining Python/FastAPI backends with React, Next.js, and React Native frontends, integrating third-party APIs, automation platforms, databases, and cloud services.
• Worked directly with international clients on architecture, technical planning, implementation, integrations, debugging, deployment, and production optimization.